You can deploy this model and start earning money today!

Drshyamsundaramindia/
finetuned-mistral-7b-instruct-v0.2

Name: finetuned-mistral-7b-instruct-v0.2
Availability: InStock

Views

No views yet

Fine-tuned mistral-7b-instruct-v0.2 Model

This model is a fine-tuned version of mistral-7b-instruct-v0.2 using unsloth and QLoRA, specifically trained on Bajaj Finserv financial earnings documents from the official https://www.aboutbajajfinserv.com/investor-relations-financial-results.

Model Description

This model has been fine-tuned to understand and generate text related to financial reports. It can be used for tasks such as summarization, question-answering, and information extraction from similar documents.

Training Data

The model was fine-tuned on a collection of financial earnings reports in PDF and TXT format, located in the data folder.

Training Hyperparameters

Base Model: unsloth/mistral-7b-instruct-v0.2
Fine-tuning Method: QLoRA via unsloth
Epochs: 3
Learning Rate: 2e-4
Max Sequence Length: 2048

Usage

To use this model for inference, you can load it directly from the Hugging Face Hub (once uploaded successfully):

python

from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Drshyamsundaramindia/finetuned-mistral-7b-instruct-v0.2")
model = AutoModelForCausalLM.from_pretrained("Drshyamsundaramindia/finetuned-mistral-7b-instruct-v0.2")

# Example usage (adjust as per your task)
input_text = "Your financial document text here."
inputs = tokenizer(input_text, return_tensors="pt")
outputs = model.generate(**inputs, max_new_tokens=100)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

Limitations and Bias

This model is trained on a specific dataset of financial documents and may not generalize well to other domains. It may also inherit biases present in the training data.

Further Information

For more details on the fine-tuning process, refer to the accompanying notebook.